Synthesis of chiral 3,4,5,6-tetrahydro-1,4-thiazin-2-ones (thiamorpholin-2-ones) - Novel heterocycles possessing enhanced carbonyl electrophilicity over their oxygen counterparts

Article abstract of DOI:10.1055/s-2006-951548

We report herein the first synthesis of chiral derivatives possessing the 1,4-thiazinone core. As predicted, the thiolactone is more susceptible to nucleophilic attack than the equivalent lactone system. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2006-951548

LETTER
3259
Synthesis of Chiral 3,4,5,6-Tetrahydro-1,4-thiazin-2-ones (Thiamorpholin-2-
ones) – Novel Heterocycles Possessing Enhanced Carbonyl Electrophilicity
over their Oxygen Counterparts
S
ynthesis of Chiral 3,4
i
, 5,6-Te
c
trahydro-1
h
a
ne
s
(Thiam
e
orpholin-2
l
-ones) G. B. Drew, Laurence M. Harwood,* Ran Yan
Department of Chemistry, University of Reading, Whiteknights, Reading, Berkshire, RG6 6AD, UK
Fax +44(118)3786121; E-mail: l.m.harwood@reading.ac.uk
Received 20 August 2006
ethyl chloroformate at 0 °C and then reacted with sodium
Abstract: We report herein the first synthesis of chiral derivatives
possessing the 1,4-thiazinone core. As predicted, the thiolactone is
more susceptible to nucleophilic attack than the equivalent lactone
system.
hydrosulfide hydrate to form a dark green solution within
30 minutes which was acidified by addition of 1 M hydro-
gen chloride to pH 2. The thioacids 2a–d thus obtained
were converted into the corresponding potassium salts
using potassium 2-ethyl hexanoate (KEH) in diethyl ether
following a previously described protocol.4 In the case of
thioproline derivative 2e, the potassium salt was generat-
ed using potassium 2,6-di(tert-butyl)-4-methylphenoxide.
These salts were then reacted with 2-bromoacetophenone
to generate the thioesters 3a–e in yields of (68%, 53%,
69%, 37% and 83% yields, respectively, over the whole
sequence (Scheme 1).
Key words: chiral, 1,4-thiazinone, thiamorpholinone
We have published extensively in the area of chirality re-
lay using chiral morpholinones as substrates for azo-
methine ylide generation and trapping,1 observing on
occasion nucleophilic ring opening of the lactone.2 Wish-
ing to utilize this side reactivity in another application, we
decided to synthesize thiamorpholinone analogues in the
expectation that nucleophilic attack on the thiolactone of
these templates (for which no general synthetic pathway
exists3) should occur more readily and under milder con-
ditions.
After removing the Boc group from 3a with 20% trifluo-
roacetic acid in dichloromethane, the cyclization was ini-
tially attempted using pH 5 acetate buffer, conditions
successful in the morpholinone series, but after 18 hours
only decomposed materials were recovered from the
aqueous phase; presumably due to hydrolysis of the
thioester. However, when the deprotected amine was
treated with anhydrous potassium carbonate in dry dichlo-
romethane, a new material was observed by TLC analysis
and the reaction was found to be complete within 24
hours.6 On work-up and purification a 59% yield of cy-
clized material was obtained and this was found to be a
mixture of two inseparable isomers in a ratio of 2:1. The
major component was shown to be the expected imine 4a
by analogy with the observations of Caplar and Sunjic and
We proposed that the desired 3,5-disubstituted thiamor-
pholinones could be constructed by conversion of Boc-
protected amino acids to their corresponding potassium
thionate salts according to literature precedent;4 then
adapting Caplar and Sunjic’s method for morpholinone
synthesis.5 However, we recognized that we could run
into difficulties at the final hydrogenation step, as the
presence of sulfur in the substrates was likely to poison
any transition-metal catalyst and it was not predictable
whether we could achieve efficient diastereocontrol with
alternative reducing agents if catalyzed hydrogenation
proved impossible.
1
the minor one was identified as the enamine 5a by H
NMR spectroscopic analysis. Two doublets at d = 4.22
(J = 15.3 Hz) ppm and d = 3.96 (J = 15.3 Hz) ppm corre-
sponded to the methylene group in the imine 4a and a
The triethylamine salts of five representative Boc-protect-
ed L-amino acids 1a–e were C-terminus-activated with
(i) potassium 2-ethylhexanoate or
potassium 2,6-di(tert-butyl-4-
methylphenoxide, Et2O, 30 min
(i) Et3N, DMF, 0 °C, 30 min
(ii) ClCO2Et, 0 °C, 30 min
R1
BocR2N
H
R1
BocR2N
H
R1
BocR2N
H
O
COSH
CO2H
(ii) 2-bromoacetophenone, DMF, 18 h
(iii) NaHS, 0 °C, 30 min
(iv) 1 M HCl, 0 °C
S
a R1 = Me, R2 = H
68%
1
2
b R1 = i-Pr, R2 = H
53%
69%
37%
83%
c R1 = Bz, R2 = H
O
Ph
d R1 = CH2CH(CH3)2, R2 = H
e R1, R2 = –(CH2)3–
3
Scheme 1
SYNLETT 2006, No. 19, pp 3259–3262
0
1
.1
2
.2
0
0
6
Advanced online publication: 23.11.2006
DOI: 10.1055/s-2006-951548; Art ID: G27506ST
© Georg Thieme Verlag Stuttgart · New York

3260
M. G. B. Drew et al.
LETTER
singlet at d = 5.22 ppm corresponded to the alkene proton ing to the ABX system at C-5 and C-6 being diagnostic;
of the enamine.7 Substrates 3b–d behaved similarly fur- the relative stereochemistry being concluded initially
nishing mixtures of cyclized materials in 49%, 53% and from NOE difference experiments. Further evidence for
57% yields, respectively, with the exception that cesium reduction was obtained from the mass spectrum in which
carbonate proved more effective for producing the 3-iso- the mass ion at 236 [MH+] could be observed.
propyl-substituted isomers 4b and 5b and the 3-benzyl
NaBH3CN, AcOH,
MeOH, 24 h
isomers 4c and 5c were obtained in a 1:1 ratio (Scheme 2).
H
H
H
N
Ph
In the case of 3e, cyclization led to 5e in an isolated crude
yield of 72%, but the product proved too labile to purify
completely.
4b, 5b
51%, >97% de
O
S
6b
With the 3,5-disubstituted dihydro-1,4-thiazin-2-ones 4
and 5 in hand, our final task was to reduce them to the de-
sired thiamorpholinones 6. However, using the mixture of
3-methyl-substituted substrates 4a and 5a our earlier fears
were realized when it was found that these compounds re-
sisted various catalytic hydrogenation conditions. More-
over, when treated with hydride donors such as sodium
borohydride or sodium cyanoborohydride in methanol,
only decomposition was observed. Frustratingly, a combi-
nation of aluminium and nickel chloride in aqueous THF8
was successful on the initial attempt (albeit giving low
yield and stereoselectivity) but could not be repeated de-
spite numerous attempts and modifications of the condi-
tions.
Scheme 3
Furthermore, from inspection of the 1H NMR spectrum of
the crude material it was clear that this reduction was
highly diastereoselective (>97% de). The conclusions
based upon spectroscopic analysis were upheld by X-ray
crystallographic analysis9 which showed that the thiamor-
pholinone ring adopts a flattened chair conformation with
the syn-3,5-substituents in equatorial environments, in di-
rect analogy with the analogous morpholinone systems
previously studied (Figure 1).10 From this we conclude
that the hydride attacks the intermediate iminium species
from the opposite face to the isopropyl group.
Although decomposition occurred under the sodium boro-
hydride or sodium cyanoborohydride conditions, the
enamine olefin proton signal in the 1H NMR spectrum and
the imine carbon signal in the 13C NMR spectrum of crude
material also disappeared. From these observations, we
proposed that the imine and the enamine had been re-
duced, but the resultant 3-methylthiamorpholinone 6a
was subsequently destroyed. Believing this might be due
to lability of the thiolactone, we turned our attention to the
3-isopropyldidehydrothiamorpholinones 4b and 5b when
the reduced product might be more resistant to subsequent
decomposition as a result of the Thorpe–Ingold effect
favouring ring closure.
Figure 1 Crystal structure of 6b
Based on previous observations, sodium cyanoborohy-
dride in methanol was chosen as the reducing reagent to
attempt the transformation of the mixture of 4b and 5b. It
was found that the mixture of starting materials was slow-
ly converted into a single less polar compound and the
addition of a few drops of acetic acid greatly accelerated
this process. The new compound was isolated in 51%
purified yield and identified as the desired syn-3(S)-iso-
However, when we turned our attention to the analogues
4c,d and 5c,d, not only did stereoselective reduction occur
but the resultant thiolactone underwent nucleophilic at-
tack by the solvent giving rise to the open-chain methyl
esters 7c,d; whereas the alanine and proline derived mate-
rials 4a, 5a,e decomposed, providing clear evidence of the
greater propensity of the thiolactone to undergo nucleo-
philic attack compared to the morpholinone analogues
which are stable towards these conditions (Scheme 4).
1
propyl-5(R)-phenylthiamorpholinone (6b) by H NMR
analysis (Scheme 3); three double doublets at d = 4.23
(J = 10.9 Hz, J¢ = 2.9 Hz) ppm, d = 3.46 (J = 11.6 Hz, J¢ =
10.9 Hz,) d = 3.07 (J = 11.6 Hz, J¢ = 2.9 Hz) correspond-
R1
BocR2N
H
R2
N
H
H
yield
59%
49%
53%
57%
72%
ratio
R1
O
N
S
R1
Ph
Ph
(i) TFA, CH2Cl2, r.t., 2 h
4
1
1
1
1
0
:
5
2
2
1
2
1
S
a
b
c
d
e
(ii) Cs2CO3 or K2CO3,
CH2Cl2, r.t., 24 h
O
O
S
O
Ph
4
5
3
Scheme 2
Synlett 2006, No. 19, 3259–3262 © Thieme Stuttgart · New York

LETTER
Synthesis of Chiral 3,4,5,6-Tetrahydro-1,4-thiazin-2-ones (Thiamorpholin-2-ones)
3261
(3a, 3.88 g, 12 mmol) in anhyd CH2Cl2 (50 mL) was added
TFA (12.5 mL). The resulting solution was stirred under an
atmosphere of nitrogen for 2 h. The solvent and the excess of
TFA were removed in vacuo and the residue was dissolved
in anhyd CH2Cl2 (50 mL) to which K2CO3 (8.28 g, 60 mmol,
5.0 equiv) was added. The resulting mixture was stirred
under an atmosphere of nitrogen for 24 h. Filtration through
a short pad of Celite® and removal of solvent from the filtrate
in vacuo gave the crude material which was purified by flash
column chromatography on silica, eluting with light PE and
Et2O (4:1) to furnish the inseparable mixture of title products
as a yellow oil (1.44 g, 59%).
NaBH3CN, AcOH,
MeOH, 24 h
R2
N
H
H
R1
Ph
4, 5
a
c
d
e
dec.
72%
61%
dec.
CO2Me
SH
7
Scheme 4
However, the simple expedient of substituting THF for
the methanol solvent permitted isolation of the desired
3,5-disubstituted thiamorpholinones 6a–e in 62%, 70%,
86%, 62% and 46% yields, respectively, with excellent
diastereoselectivity (>97% from examination of crude
product mixtures) in all cases (Scheme 5).
Reduction of 4a and 5a to Produce 6a.
To a solution of (S)-3-methyl-5-phenyl-4,5-didehydro-1,4-
thiazin-2-one (4a) and (S)-3-methyl-5-phenyl-5,6-
didehydro-1,4-thiazin-2-one (5a, 100 mg, 0.49 mmol) in
anhyd THF (5 mL) was added sodium cyanoborohydride (62
mg, 0.97 mmol 2.0 equiv) and AcOH (ca. 5 drops). The
resulting solution was stirred under an atmosphere of
nitrogen for 12 h. The solvent was removed in vacuo to give
the crude material which was purified by flash column
chromatography on silica, eluting with light PE and Et2O
(8:1, then 4:1) to furnish the title product as a pale yellow oil
(63 mg, 62%).
Reduction and Thiolactone Cleavage to Produce 7d.
To a solution of (S)-3-(2-methylpropyl)-5-phenyl-4,5-
didehydro-1,4-thiazin-2-one (4d) and (S)-3-(2-methyl-
propyl)-5-phenyl-5,6-didehydro-1,4-thiazin-2-one (5d, 92
mg, 0.38 mmol) in anhyd MeOH (8 mL) was added sodium
cyanoborohydride (70 mg, 1.12 mmol, 3.0 equiv) and AcOH
(ca. 5 drops). The resulting solution was stirred under an
atmosphere of nitrogen for 24 h. The solvent was removed in
vacuo to give the crude material which was purified by flash
column chromatography on silica, eluting with light PE and
Et2O (4:1) to furnish the title product as a pale yellow oil
(56 mg, 61%).
R2
N
H
H
NaBH3CN, AcOH,
THF, 12 h
R1
Ph
4, 5
> 97% de
O
S
a
b
c
d
e
62%
70%
86%
62%
46%
6
Scheme 5
In conclusion, we have developed an efficient synthetic
route to novel chiral thiamorpholinones, which demon-
strate increased propensity towards nucleophilic ring
opening compared with their morpholinone counterparts.
Acknowledgment
(7) Representative Analytical Data.
Compounds 4a and 5a (mixture): IR (film): nmax = 3391 (N–
H), 2983 (C–H), 1683 (C=O, thiolactone) cm–1. 1H NMR
(250 MHz, CDCl3) d = 7.74–7.27 (5 H, m, Ph), 5.22 (0.35 H,
s, PhC=CH), 4.22 (0.65 H, d, J = 15.3 Hz, CH2S), 4.02–3.97
(0.65 H, m, CH3CH), 3.96 (0.65 H, d, J = 15.3 Hz, CH2S),
3.66–3.61 (0.35 H, m, CH3CH), 1.59 (1.95 H, d, J = 6.3 Hz,
CHCH3), 1.32 (1.05 H, d, J = 7.3 Hz, CHCH3). MS (CI):
m/z (%) = 205 (44) [M+], 181 (66), 162 (100). HRMS: m/z
calcd for C11H11NOS: 205.0561; found: 205.0562.
Compound 6a: IR (film): nmax = 3319 (N–H), 2929 (C–H),
1668 (C=O, thiolactone) cm–1. 1H NMR (250 MHz, CDCl3):
d = 7.45–7.32 (5 H, m, Ph), 4.30 (1 H, dd, J = 10.82 Hz, J¢ =
3.31 Hz, PhCH), 3.86 (1 H, q, J = 6.67 Hz, CHCH3), 3.55 (1
H, dd, J = 11.69 Hz, J¢ = 10.82 Hz, CH2S), 3.13 (1 H, dd,
J = 11.69 Hz, J¢ = 3.31 Hz, 1 × CH2S), 1.89 (1 H, br, NH),
1.40 (1 H, d, J = 6.67 Hz, CHCH3). 13C NMR (62.5 MHz,
CDCl3): d = 200.6, 142.3, 129.4, 128.8, 126.8, 65.2, 60.1,
38.8, 18.2. MS (CI): m/z (%) = 208 (51) [MH+], 131 (46),
121 (100). HRMS: m/z calcd for C11H14NOS: 208.0796;
found: 208.0788. [a]43620 –1.0 (c 0.70 CHCl3), [a]D20 0.0 (c
0.70 CHCl3).
Compound 7d: IR (film): nmax = 3308 (N–H), 2956 (C–H),
1734 (C=O, ester) cm–1. 1HNMR (250 MHz, CDCl3): d =
7.29–7.17 (5 H, m, Ph), 3.63 (3 H, s, CH3O), 3.47 (1 H, dd,
J = 8.7 Hz, J¢ = 4.8 Hz, PhCH), 2.95 [1 H, dd, J = 8.9 Hz,
J¢ = 5.5 Hz, (CH3)2CHCH2CH], 2.76 (1 H, dd, J = 13.3 Hz,
J¢ = 4.6 Hz, CH2S), 2.51 (1 H, m, CH2S), 1.78–1.72 [1 H, m,
(CH3)2CHCH2], 1.41–1.26 [2 H, m, (CH3)2CHCH2], 0.79 [3
H, d, J = 6.6 Hz, 3 × (CH3)2CH], 0.62 [3 H, d, J = 6.6 Hz, 3
× (CH3)2CH]. 13C NMR (62.5 MHz, CDCl3): d = 175.5,
We thank the Overseas Research Students Award Scheme and The
University of Reading for financial support to R.Y.
References and Notes
(1) For recent communications, see: (a) Draffin, W. N.;
Harwood, L. M. Synlett 2006, 857. (b) Aldous, D. J.; Drew,
M. G. B.; Draffin, W. N.; Hamelin, E. M.-N.; Harwood, L.
M.; Thurairatnam, S. Synthesis 2005, 3271. (c) Aldous, D.
J.; Hamelin, E. M.-N.; Harwood, L. M.; Thurairatnam, S.
Synlett 2001, 1841. (d) Aldous, D. J.; Drew, M. G. B.;
Hamelin, E. M.-N.; Harwood, L. M.; Jahans, A. B.;
Thurairatnam, S. Synlett 2001, 1836; see also references
cited therein.
(2) (a) Harwood, L. M.; Robertson, S. M. J. Chem. Soc., Chem.
Commun. 1998, 2641. (b) Alker, D.; Hamblett, G.;
Harwood, L. M.; Robertson, S. M.; Watkin, D. J.; Williams,
C. E. Tetrahedron 1998, 54, 6089.
(3) A few reports of thiamorpholinones embedded within b-
lactam structures have appeared: (a) Costerousse, G.;
Cagniant, A.; Didierlaurent, S. Bull. Soc. Chim. Fr. 1989,
830. (b) Hakimelahi, G. H.; Tsay, S.-C.; Ramezani, Z.; Hwu,
J. R. Helv. Chim. Acta 1996, 79, 813.
(4) Gottstein, W. J.; Babel, R. B.; Crast, L. B. J. Med. Chem.
794, 8, 1965.
(5) Caplar, V.; Lisini, A.; Kajfez, F.; Sunjic, V. J. Org. Chem.
1978, 43, 1355.
(6) Representative Experimental Procedures.
Cyclization of 3a to Produce 4a and 5a.
To a solution of 2-oxo-2-phenylethyl-(S)-N-Boc-thioalanate
Synlett 2006, No. 19, 3259–3262 © Thieme Stuttgart · New York

3262
M. G. B. Drew et al.
LETTER
140.8, 127.5, 126.8, 126.5, 62.6, 55.9, 50.6, 42.2, 31.8, 23.6,
22.1, 20.7. MS (CI): m/z (%) = 282 (55) [M+], 234 (100), 174
(16). HRMS: m/z calcd for C15H24NO2S: 282.1528; found:
282.1526. [a]D20 –4.6 (c 0.77 CHCl3).
bond lengths and angles and thermal parameters have been
deposited with the Cambridge Crystallographic Database
Service. Please quote reference number CCDC 617867.
(10) For example, see: (a) Drew, M. G. B.; Harwood, L. M.;
Price, D. W.; Park, G. Tetrahedron Lett. 2000, 5077.
(b) Drew, M. G. B.; Harwood, L. M.; Park, G.; Price, D. W.;
Tyler, S. N. G.; Park, C. R.; Gho, S. G. Tetrahedron 2001,
57, 5641.
(11) Kabsch, W. J. Appl. Cryst. 1988, 21, 916.
(12) Sheldrick, G. M. Acta Crystallogr., Sect. A: Fundam.
Crystallogr. 1990, 46, 467.
(13) Sheldrick, G. M. Program for Crystal Structure Refinement;
University of Göttingen: Germany, 1993.
(8) Bhabani, K.; Barua, N. C. Tetrahedron 1991, 47, 8587.
(9) X-ray Crystal Data of 6b.
C13H17NOS, M = 235.34, triclinic, space group P1, Z = 8,
a =& nbsp;9.7241 (8), b = 13.4169 (12), c = 19.8594 (17),
V = 2437 Å3, d = 1.283 g cm3, 12166 independent
reflections were collected with MoKa radiation on a
MARresearch Image Plate system. Data analysis was carried
out with the XDS program,11 the structure was solved by
direct methods using Shelx8612 and refined on F2 using
Shelx13 to R1 = 0.1039, wR2 = 0.1641. Atomic coordinates,
Synlett 2006, No. 19, 3259–3262 © Thieme Stuttgart · New York